CSCI 5582 Fall 2006 CSCI 5582 Artificial Intelligence Lecture 24 Jim Martin.
Natural Language Processing Lecture 6—9/17/2013 Jim Martin.
-
Upload
roland-mcgee -
Category
Documents
-
view
213 -
download
0
Transcript of Natural Language Processing Lecture 6—9/17/2013 Jim Martin.
Natural Language Processing
Lecture 6—9/17/2013
Jim Martin
04/21/23 Speech and Language Processing - Jurafsky and Martin 2
Today
More language modeling with N-grams Basic counting Probabilistic model
Independence assumptions
Simple smoothing methods
04/21/23 Speech and Language Processing - Jurafsky and Martin 3
N-Gram Models
We can use knowledge of the counts of N-grams to assess the conditional probability of candidate words as the next word in a sequence.
Or, we can use them to assess the probability of an entire sequence of words. Pretty much the same thing as we’ll
see...
04/21/23 Speech and Language Processing - Jurafsky and Martin 4
Language Modeling
Back to word prediction We can model the word prediction
task as the ability to assess the conditional probability of a word given the previous words in the sequence P(wn|w1,w2…wn-1)
We’ll call a statistical model that can assess this a Language Model
04/21/23 Speech and Language Processing - Jurafsky and Martin 5
Language Modeling
How might we go about calculating such a conditional probability? One way is to use the definition of
conditional probabilities and look for counts. So to get
P(the | its water is so transparent that) By definition that’s
P(its water is so transparent that the) P(its water is so transparent that)
04/21/23 Speech and Language Processing - Jurafsky and Martin 6
Very Easy Estimate
How to estimate? P(the | its water is so transparent that)
P(the | its water is so transparent that) =Count(its water is so transparent that the) Count(its water is so transparent that)
04/21/23 Speech and Language Processing - Jurafsky and Martin 7
Language Modeling
Unfortunately, for most sequences and for most text collections we won’t get good estimates from this method. What we’re likely to get is 0. Or worse 0/0.
Clearly, we’ll have to be a little more clever. Let’s first use the chain rule of probability And then apply a particularly useful
independence assumption
04/21/23 Speech and Language Processing - Jurafsky and Martin 8
The Chain Rule
Recall the definition of conditional probabilities
Rewriting:
For sequences... P(A,B,C,D) = P(A)P(B|A)P(C|A,B)P(D|A,B,C)
In general P(x1,x2,x3,…xn) = P(x1)P(x2|x1)P(x3|x1,x2)…P(xn|
x1…xn-1)
04/21/23 Speech and Language Processing - Jurafsky and Martin 9
The Chain Rule
P(its water was so transparent)=P(its)* P(water|its)* P(was|its water)* P(so|its water was)* P(transparent|its water was so)
04/21/23 Speech and Language Processing - Jurafsky and Martin 10
Unfortunately
That doesn’t really help since it relies on having N-gram counts for a sequence that’s only 1 shorter than what we started with Not likely to help with getting counts
In general, we will never be able to get enough data to compute the statistics for those longer prefixes Same problem we had for the strings
themselves
04/21/23 Speech and Language Processing - Jurafsky and Martin 11
Independence Assumption
Make a simplifying assumption P(lizard|
the,other,day,I,was,walking,along,and,saw,a) = P(lizard|a)
Or maybe P(lizard|
the,other,day,I,was,walking,along,and,saw,a) = P(lizard|saw,a)
That is, the probability in question is to some degree independent of its earlier history.
04/21/23 Speech and Language Processing - Jurafsky and Martin 12
Independence Assumption
This particular kind of independence assumption is called a Markov assumption after the Russian mathematician Andrei Markov.
04/21/23 Speech and Language Processing - Jurafsky and Martin 13
Replace each component in the product with a shorter approximation (assuming a prefix of N - 1)
Bigram (N=2) version
Markov Assumption
Bigram ExampleP(its water was so transparent)=P(its)* P(water|its)* P(was|its water)* P(so|its water was)* P(transparent|its water was so)
P(its water was so transparent)=
P(its)* P(water|its)* P(was|water)* P(so|was)* P(transparent|so)
04/21/23 Speech and Language Processing - Jurafsky and Martin 14
04/21/23 Speech and Language Processing - Jurafsky and Martin 15
Estimating Bigram Probabilities
The Maximum Likelihood Estimate (MLE)
04/21/23 Speech and Language Processing - Jurafsky and Martin 16
An Example
<s> I am Sam </s> <s> Sam I am </s> <s> I do not like green eggs and ham </s>
04/21/23 Speech and Language Processing - Jurafsky and Martin 17
Maximum Likelihood Estimates
The maximum likelihood estimate of some parameter of a model M from a training set T Is the estimate that maximizes the likelihood of the
training set T given the model M Suppose the word “Chinese” occurs 400 times in a
corpus of a million words (Brown corpus) What is the probability that a random word from
some other text from the same distribution will be “Chinese”
MLE estimate is 400/1000000 = .004 This may be a bad estimate for some other corpus
But it is the estimate that makes it most likely that “Chinese” will occur 400 times in a million word corpus.
04/21/23 Speech and Language Processing - Jurafsky and Martin 18
Berkeley Restaurant Project Sentences
can you tell me about any good cantonese restaurants close by
mid priced thai food is what i’m looking for tell me about chez panisse can you give me a listing of the kinds of food that
are available i’m looking for a good place to eat breakfast when is caffe venezia open during the day
04/21/23 Speech and Language Processing - Jurafsky and Martin 19
Bigram Counts
Out of 9222 sentences Eg. “I want” occurred 827 times
04/21/23 Speech and Language Processing - Jurafsky and Martin 20
Bigram Probabilities
Divide bigram counts by prefix unigram counts to get probabilities.
04/21/23 Speech and Language Processing - Jurafsky and Martin 21
Bigram Estimates of Sentence Probabilities
P(<s> I want english food </s>) = P(i|<s>)*
P(want|I)* P(english|want)* P(food|english)* P(</s>|food)* =.000031
04/21/23 Speech and Language Processing - Jurafsky and Martin 22
Kinds of Knowledge
P(english|want) = .0011
P(chinese|want) = .0065
P(to|want) = .66 P(eat | to) = .28 P(food | to) = 0 P(want | spend) = 0 P (i | <s>) = .25
As crude as they are, N-gram probabilities capture a range of interesting facts about language.
World knowledge
Syntax
Discourse
04/21/23 Speech and Language Processing - Jurafsky and Martin 23
Shannon’s Game
Assigning probabilities to sentences is all well and good, but it’s not terribly entertaining. What if we turn these models around and use them to generate random sentences that are like the sentences from which the model was derived.
04/21/23 Speech and Language Processing - Jurafsky and Martin 24
Shannon’s Method Sample a random bigram (<s>, w) according to its probability Now sample a random bigram (w, x) according to its
probability Where the prefix w matches the suffix from the first choice.
And so on until we randomly choose a (y, </s>). Stop. Then string the words together <s> I I want
want to to eat eat Chinese
Chinese food food </s>
If the model is good, then the generated strings should be pretty reasonable.
04/21/23 Speech and Language Processing - Jurafsky and Martin 25
Shakespeare
04/21/23 Speech and Language Processing - Jurafsky and Martin 26
Shakespeare as a Corpus
N=884,647 tokens, V=29,066 Shakespeare produced 300,000 bigram
types out of V2= 844 million possible bigrams... So, 99.96% of the possible bigrams were
never seen (have zero entries in the table) This is the biggest problem in language
modeling; we’ll come back to it.
4-grams are worse... What's coming out looks like Shakespeare because it is Shakespeare
04/21/23 Speech and Language Processing - Jurafsky and Martin 27
Break
News Get the assignment in by Friday
I’ll post a test set later this week
First quiz (Chapters 1 to 6) pushed back to October 3 Quiz is based on the readings and what we do in
class. I will give specific readings for each chapter
And I’ll post some past quizzes on this material
04/21/23 Speech and Language Processing - Jurafsky and Martin 28
Model Evaluation
How do we know if our models are any good? And in particular, how do we know if one model
is better than another. Shannon’s game gives us an intuition.
The generated texts from the higher order models sure look better. That is, they sound more like the text the model was
obtained from. The generated texts from the various
Shakespeare models look different That is, they look like they’re based on different
underlying models.
But what does that mean? Can we make that notion operational?
04/21/23 Speech and Language Processing - Jurafsky and Martin 29
Evaluation Standard method
Train parameters of our model on a training set.
Look at the models performance on some new data This is exactly what happens in the real world; we
want to know how our model performs on data we haven’t seen
So use a test set. A dataset which is different than our training set, but is drawn from the same source
Then we need an evaluation metric to tell us how well our model is doing on the test set. One such metric is perplexity
04/21/23 Speech and Language Processing - Jurafsky and Martin 30
Perplexity
The intuition behind perplexity as a measure is the notion of surprise. How surprised is the language model
when it sees the test set? Where surprise is a measure of...
Gee, I didn’t see that coming...
The more surprised the model is, the lower the probability it assigned to the test set
The higher the probability, the less surprised it was
04/21/23 Speech and Language Processing - Jurafsky and Martin 31
Perplexity
Perplexity is the probability of a test set (assigned by the language model), normalized by the number of words:
Chain rule:
For bigrams:
Minimizing perplexity is the same as maximizing probability The best language model is one that best
predicts an unseen test set
04/21/23 Speech and Language Processing - Jurafsky and Martin 32
Lower perplexity means a better model
Training 38 million words, test 1.5 million words, WSJ
04/21/23 Speech and Language Processing - Jurafsky and Martin 33
Practical Issues
We do everything in log space Avoid underflow Also adding is faster than multiplying
04/21/23 Speech and Language Processing - Jurafsky and Martin 34
Unknown Words
In the training data, the vocabulary is what’s there. But once we start looking at test data, we will run into words that we haven’t seen in training With an Open Vocabulary task
Create an unknown word token <UNK> Create a fixed lexicon L, of size V
From a dictionary or A subset of terms from the training set
At text normalization phase, any training word not in L changed to <UNK>
Now we count that like a normal word At test time
Use UNK counts for any word not in training
04/21/23 Speech and Language Processing - Jurafsky and Martin 35
SmoothingDealing w/ Zero Counts
Back to Shakespeare Recall that Shakespeare produced 300,000
bigram types out of V2= 844 million possible bigrams...
So, 99.96% of the possible bigrams were never seen (have zero entries in the table)
Does that mean that any sentence that contains one of those bigrams should have a probability of 0?
For generation (shannon game) it means we’ll never emit those bigrams
But for analysis it’s problematic because if we run across a new bigram in the future then we have no choice but to assign it a probability of zero.
04/21/23 Speech and Language Processing - Jurafsky and Martin 36
Zero Counts Some of those zeros are really zeros...
Things that really aren’t ever going to happen On the other hand, some of them are just rare events.
If the training corpus had been a little bigger they would have had a count
What would that count be in all likelihood? Zipf’s Law (long tail phenomenon):
A small number of events occur with high frequency A large number of events occur with low frequency You can quickly collect statistics on the high frequency events You might have to wait an arbitrarily long time to get valid
statistics on low frequency events Result:
Our estimates are sparse! We have no counts at all for the vast number of things we want to estimate!
Answer: Estimate the likelihood of unseen (zero count) N-grams!
04/21/23 Speech and Language Processing - Jurafsky and Martin 37
Laplace Smoothing
Also called Add-One smoothing Just add one to all the counts! Very simple
MLE estimate:
Laplace estimate:
04/21/23 Speech and Language Processing - Jurafsky and Martin 38
Bigram Counts
04/21/23 Speech and Language Processing - Jurafsky and Martin 39
Laplace-Smoothed Bigram Counts
04/21/23 Speech and Language Processing - Jurafsky and Martin 40
Laplace-Smoothed Bigram Probabilities
04/21/23 Speech and Language Processing - Jurafsky and Martin 41
Reconstituted Counts
Reconstituted Counts (2)
04/21/23 Speech and Language Processing - Jurafsky and Martin 42
04/21/23 Speech and Language Processing - Jurafsky and Martin 43
Big Change to the Counts!
C(want to) went from 608 to 238! P(to|want) from .66 to .26! Discount d= c*/c
d for “chinese food” =.10!!! A 10x reduction So in general, Laplace is a blunt instrument Could use more fine-grained method (add-k)
Because of this Laplace smoothing not often used for language models, as we have much better methods
Despite its flaws Laplace (add-1) is still used to smooth other probabilistic models in NLP and IR, especially For pilot studies In document classification In domains where the number of zeros isn’t so huge.
Foreshadowing
Just a hint of some stuff we’ll get to later
With Add-1 we counted, noted all the zeros, then added 1 to everything to get our new counts
What if we did the Add-1 before we did any counting What would that model look like?
What kind of distribution is it?
What happens to that model as we start doing our counting?
04/21/23 Speech and Language Processing - Jurafsky and Martin 44
04/21/23 Speech and Language Processing - Jurafsky and Martin 45
Better Smoothing
An intuition used by many smoothing algorithms Good-Turing Kneser-Ney Witten-Bell
Is to use the count of things we’ve seen once to help estimate the count of things we’ve never seen
04/21/23 Speech and Language Processing - Jurafsky and Martin 46
Types, Tokens and Squirrels
Much of what’s coming up was first studied by field biologists who are often faced with 2 related problems Determining how many species occupy
a particular area (types) And determining how many individuals
of a given species are living in a given area (tokens)
04/21/23 Speech and Language Processing - Jurafsky and Martin 47
Good-Turing Josh Goodman Intuition
Imagine you are fishing You know there are 8 species: carp, perch, whitefish, trout,
salmon, eel, catfish, bass Not exactly sure where such a situation would arise...
You have caught up to now 10 carp, 3 perch, 2 whitefish, 1 trout, 1 salmon, 1 eel = 18
fish How likely is it that the next fish to be caught is an
eel?
Slide adapted from Josh Goodman
How likely is it that the next fish caught will be a member of previously uncaptured species?
Now how likely is it that the next fish caught will be an eel?
04/21/23 Speech and Language Processing - Jurafsky and Martin 48
Notation: Nx is the frequency-of-frequency-x So N10=1
Number of fish species seen 10 times is 1 (carp) N1=3
Number of fish species seen once is 3 (trout, salmon, eel)
To estimate total number of unseen species Use number of species seen once c0
* =c1 p0 = N1/N All other estimates (counts) are adjusted
downward to account for unseen probabilities
Good-Turing
Slide from Josh Goodman
Eel = c*(1) = (1+1) 1/ 3 = 2/3
04/21/23 Speech and Language Processing - Jurafsky and Martin 49
GT Fish Example
04/21/23 Speech and Language Processing - Jurafsky and Martin 50
GT Smoothed Bigram Probabilities
04/21/23 Speech and Language Processing - Jurafsky and Martin 51
In practice, assume large counts (c>k for some k) are reliable:
Also: we assume singleton counts c=1 are unreliable, so treat N-grams with count of 1 as if they were count=0
Also, need each N_k to be non-zero, so we need to smooth (interpolate) the N_k counts before computing c* from them
GT Complications
04/21/23 Speech and Language Processing - Jurafsky and Martin 52
Next Time
On to Chapter 5 Parts of speech Part of speech tagging and HMMs